[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: really blockers for (an edu0) release?

[Holger Levsen]
> Hi,
> http://wiki.debian.org/DebianEdu/Status/Wheezy lists these blockers
> for a release:
> 1. KDE applications (at least konqueror and step) do not use the
> proxy by default, and automatic proxy configuration do not work,
> making these tools useless on a network where proxy use is
> required. Looks like KDE do not understand our wpad.dat file?

I suspect a fix for this require a major update of KDE.  Perhaps we
can document the problem and come up with a wpad.dat file that KDE
understand for those environments?

No idea if it was like this also in Squeeze.  Is it a regression or

> 2. Chromium fail to work when it is set to use automatic proxy
> configuration and no wpad.dat is found on a network where proxy use
> is required, making it some times fail to work on standalone and
> roaming workstations when the machines are moved outside the
> backbone network.

Also nasty.  Might be connected to the KDE problem, as Chromium fetch
the proxy setup from KDE when logged into KDE.

We did not include Chromium in Squeeze, so this isn't a regression.

> 3. Iceweasel do not update its default start page from LDAP at boot
> like it did in Squeeze. This make it impossible for the school admin
> to centrally set the default browser start page.

This is a regression and breaking a adverticed feature.  I believe it
is vital to fix it before release.  I commited a fix a few minutes

> IMO these are more than just "known problems" but rather "issues
> which should be fixed", but definitly not blockers for a release.

Well, I believe we should at least try to solve these, as they will
block Debian Edu from providing a working solution for several schools
here in Norway.

> Do you agree? Do we have any actual blockers? Except source dvd and,
> thats it?!

I also have seen very slow boot with LTSP and slow login with only one
client.  This might be important to figure out and fix to make sure a
complete class room can boot LTSP clients.  Anyone able to test how
long it take to boot 30 LTSP clients?  I lack the test bench for that.

I tried to use bootchart2 with LTSP, but the boot did not work as it
should.  Anyone been able to benchmark the LTSP boot lately?  Is it
possible to beat bootchart or bootchart2 into shape for this task?

Happy hacking
Petter Reinholdtsen

Reply to: